[libvirt] [PATCH] udev: Fix build on older platforms
Caused by commit @d1eea6c1 due to the missing symbol on older platforms. Signed-off-by: Erik Skultety <eskultet@redhat.com> --- Despite falling under build-breaker category, I'd like to get a proper review, since I'm not really familiar with autoconf and there might be a better fix. Erik configure.ac | 2 +- src/node_device/node_device_udev.c | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/configure.ac b/configure.ac index 246f4e077..b78c8b790 100644 --- a/configure.ac +++ b/configure.ac @@ -322,7 +322,7 @@ AC_CHECK_FUNCS_ONCE([cfmakeraw fallocate geteuid getgid getgrnam_r \ getmntent_r getpwuid_r getrlimit getuid if_indextoname kill mmap \ newlocale posix_fallocate posix_memalign prlimit regexec \ sched_getaffinity setgroups setns setrlimit symlink sysctlbyname \ - getifaddrs sched_setscheduler unshare]) + getifaddrs sched_setscheduler unshare udev_monitor_set_receive_buffer_size]) dnl Availability of various common headers (non-fatal if missing). AC_CHECK_HEADERS([pwd.h regex.h sys/un.h \ diff --git a/src/node_device/node_device_udev.c b/src/node_device/node_device_udev.c index a69dc1175..01438ea17 100644 --- a/src/node_device/node_device_udev.c +++ b/src/node_device/node_device_udev.c @@ -1764,12 +1764,14 @@ static int nodeStateInitialize(bool privileged, udev_monitor_enable_receiving(priv->udev_monitor); +#if HAVE_UDEV_MONITOR_SET_RECEIVE_BUFFER_SIZE /* mimic udevd's behaviour and override the systems rmem_max limit in case * there's a significant number of device 'add' events */ if (geteuid() == 0) udev_monitor_set_receive_buffer_size(priv->udev_monitor, 128 * 1024 * 1024); +#endif /* We register the monitor with the event callback so we are * notified by udev of device changes before we enumerate existing -- 2.13.0
On Wed, May 31, 2017 at 12:22:01PM +0200, Erik Skultety wrote:
Caused by commit @d1eea6c1 due to the missing symbol on older platforms.
Signed-off-by: Erik Skultety <eskultet@redhat.com> --- Despite falling under build-breaker category, I'd like to get a proper review, since I'm not really familiar with autoconf and there might be a better fix.
Erik
configure.ac | 2 +- src/node_device/node_device_udev.c | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac index 246f4e077..b78c8b790 100644 --- a/configure.ac +++ b/configure.ac @@ -322,7 +322,7 @@ AC_CHECK_FUNCS_ONCE([cfmakeraw fallocate geteuid getgid getgrnam_r \ getmntent_r getpwuid_r getrlimit getuid if_indextoname kill mmap \ newlocale posix_fallocate posix_memalign prlimit regexec \ sched_getaffinity setgroups setns setrlimit symlink sysctlbyname \ - getifaddrs sched_setscheduler unshare]) + getifaddrs sched_setscheduler unshare udev_monitor_set_receive_buffer_size])
That is always going to fail - this AC_CHECK_FUNCS_ONCE macro is only able to detect functions that are part of glibc. Take a look at m4/virt-dbus.m4 for an example of how to check for a function in another library, after detecting the library with pkgconfig
dnl Availability of various common headers (non-fatal if missing). AC_CHECK_HEADERS([pwd.h regex.h sys/un.h \ diff --git a/src/node_device/node_device_udev.c b/src/node_device/node_device_udev.c index a69dc1175..01438ea17 100644 --- a/src/node_device/node_device_udev.c +++ b/src/node_device/node_device_udev.c @@ -1764,12 +1764,14 @@ static int nodeStateInitialize(bool privileged,
udev_monitor_enable_receiving(priv->udev_monitor);
+#if HAVE_UDEV_MONITOR_SET_RECEIVE_BUFFER_SIZE /* mimic udevd's behaviour and override the systems rmem_max limit in case * there's a significant number of device 'add' events */ if (geteuid() == 0) udev_monitor_set_receive_buffer_size(priv->udev_monitor, 128 * 1024 * 1024); +#endif
/* We register the monitor with the event callback so we are * notified by udev of device changes before we enumerate existing -- 2.13.0
-- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
On 05/31/2017 12:22 PM, Erik Skultety wrote:
Caused by commit @d1eea6c1 due to the missing symbol on older platforms.
Signed-off-by: Erik Skultety <eskultet@redhat.com> --- Despite falling under build-breaker category, I'd like to get a proper review, since I'm not really familiar with autoconf and there might be a better fix.
Erik
configure.ac | 2 +- src/node_device/node_device_udev.c | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/m4/virt-udev.m4 b/m4/virt-udev.m4 index 85ca2cb1a..be7dba5d2 100644 --- a/m4/virt-udev.m4 +++ b/m4/virt-udev.m4 @@ -34,6 +34,14 @@ AC_DEFUN([LIBVIRT_CHECK_UDEV],[ if test "$with_udev_logging" = "yes" ; then AC_DEFINE_UNQUOTED([HAVE_UDEV_LOGGING], 1, [whether libudev logging can be used]) fi + + old_CFLAGS="$CFLAGS" + old_LIBS="$LIBS" + CFLAGS="$CFLAGS $UDEV_CFLAGS" + LIBS="$CFLAGS $UDEV_LIBS" + AC_CHECK_FUNCS([udev_monitor_set_receive_buffer_size]) + CFLAGS="$old_CFLAGS" + LIBS="$old_LIBS" fi ]) This is what I wanted to say, but Daniel beat me to it. ACK if you go this way. Michal
On Wed, May 31, 2017 at 12:44:37PM +0200, Michal Privoznik wrote:
On 05/31/2017 12:22 PM, Erik Skultety wrote:
Caused by commit @d1eea6c1 due to the missing symbol on older platforms.
Signed-off-by: Erik Skultety <eskultet@redhat.com> --- Despite falling under build-breaker category, I'd like to get a proper review, since I'm not really familiar with autoconf and there might be a better fix.
Erik
configure.ac | 2 +- src/node_device/node_device_udev.c | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/m4/virt-udev.m4 b/m4/virt-udev.m4 index 85ca2cb1a..be7dba5d2 100644 --- a/m4/virt-udev.m4 +++ b/m4/virt-udev.m4 @@ -34,6 +34,14 @@ AC_DEFUN([LIBVIRT_CHECK_UDEV],[ if test "$with_udev_logging" = "yes" ; then AC_DEFINE_UNQUOTED([HAVE_UDEV_LOGGING], 1, [whether libudev logging can be used]) fi + + old_CFLAGS="$CFLAGS" + old_LIBS="$LIBS" + CFLAGS="$CFLAGS $UDEV_CFLAGS" + LIBS="$CFLAGS $UDEV_LIBS" + AC_CHECK_FUNCS([udev_monitor_set_receive_buffer_size]) + CFLAGS="$old_CFLAGS" + LIBS="$old_LIBS" fi ])
This is what I wanted to say, but Daniel beat me to it. ACK if you go this way.
Thanks, help's appreciated, the fix is not pushed. Erik
participants (3)
-
Daniel P. Berrange -
Erik Skultety -
Michal Privoznik